Personal Primer is an AI-supported “Bildungsinstrument” aiming to harness advantages and possibilities offered by digital technologies for the purpose of education of elementary school pupils, most notably the acquisition of basic literacy (i.e. reading, writing, arithmetic) and promotion of informatic thinking. Currently, we have four prototypes satisfying 11 out of the 23 roadmap properties: embooked, unique, adaptable, modular, circadian, habit-disrupting, multimodal, speech-based, narrative, online-offline and monotasking. A concept of "human-machine peer learning" (HMPL) where human learns something from the machine (e.g. reading, foreign language) in the same time as machine learns something from the human (e.g. speech recognition model adaptation) is deployed in do-it-Yourself, book-like artefact narrating not only fables of ancient India.

Videos: 
https://stream.udk-berlin.de/w/gxoTarC9PK3wdumKJ8gDzx (Prix Ars Electronica 2024 submission)
https://kastalia.medienhaus.udk-berlin.de/community/data/13068.fibel_interspeech_subtitle.mp4 (accepted InterSpeech 2023 submission)